Wikipedia talk:WikiProject Military history/Assessment/Tag & Assess 2008/archive1

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

Worklists etc

Maralia has suggested 25 articles per range, which give us a navigation/tally sheet something like this:

Range adopted	Editor's name	25 done	50 done	75 done	100 done
1–99	Various editors	Done	Done	Done	Done
100–199	RichyBoy	Done	Done	Done	Done

This is fine for a 4,500 article drive but probably too long for a bigger drive.

Thoughts? --ROGER DAVIES ^talk 05:22, 6 March 2008 (UTC)

OHOH, so you just wanted to bring me here to blow holes in my ship! :)

Some random flu-and-fever-affected brainstorming:

• Can we double the width? The last tally sheet was 500 across.
•  Done is perhaps overkill - Y alone would suffice.
• Is there a compelling reason for restricting it to one tally sheet, for that matter? That bloody thing takes forever to load.
• Now that I mention the load time, a lot of that is probably due to the done templates; perhaps we should do away with a graphic template since it's apparently not at all efficient, even when using the same one repeatedly.

Of course, once my brilliant scheme to cut down on false positives is employed, 4,500 will be plenty of room, anyway, so this is all a moot point (cough). Maralia (talk) 06:33, 6 March 2008 (UTC)

• Hmmm. The last one held 170,000 articles. Even with false positives, I doubt if we'd manage to do much better than halve that.

• Overall, I think we should go 1000 articles across in four divisions. I really don't think ranges of 100 are practical. If we drop the graphics entirely, these should be much faster loading.
• There was considerable support for 1000 article ranges in the workshop. The main thing though is that the data the editor can load should correspond exactly to the range signed up for. This caused some confusion last year, with people editing other people's ranges etc.
• Agree about dropping graphic templates.
• Concerned about multiple worklists, makes it difficult to track, unless we have a starter
• Sorry, done in haste, real-life busy today.
• --ROGER DAVIES ^talk 17:55, 6 March 2008 (UTC)

I would be quite surprised if we had less than ~30,000 articles to deal with, even after the obvious false positives were filtered out. Kirill 13:54, 6 March 2008 (UTC)

• Agree. Is there any way of filtering out articles already tagged to Milhist, by the way? --ROGER DAVIES ^talk 17:55, 6 March 2008 (UTC)

We might want to split the table itself into separate sub-sections (or sub-pages, even?) for each thousand/five thousand/ten thousand/? articles; otherwise, editing a single huge table is going to become increasingly difficult. Kirill 13:54, 6 March 2008 (UTC)

No more difficult than last year, if we go for the 250 / 500/ 750 / 1000 divisions. That'll produce a shorter wider table. Thought: If the fields are partially completed in advance: "Name", "1", "2", "3", "4" it'll make navigation easier. --ROGER DAVIES ^talk 17:55, 6 March 2008 (UTC)

I didn't suspect my facetious comments about the size of the article pool might be taken seriously by anyone!

Consider the size-of-workpage issue not only in terms of my load time complaint, but also from a motivational standpoint. We give lots of barnstars for motivation - it's been suggested for as few as 20 tags - yet under the 250-by-4 format, you have to tag over 12 times that many articles just to get the satisfaction of your first measly {{done}}, and 50 times that many to reach the gratification of completing a workpage. I'm not going to keep pushing my 25-by-4 suggestion in the face of opposition, but I still think there's a value in considering smaller workpages. Maralia (talk) 07:01, 7 March 2008 (UTC)

I'd didn't adequately develop my "starter" worklist idea yesterday. In the last T&A, 20% of the editors did 80% of the tagging while many of the 80% of editors were embarking on their first real experience of wiki-gnoming. I'd expect a similar pattern this time round and I think it's crucial that we address the differing needs of both groups.

The solution is having two worklists. One, a "starter" with small ranges for people who've never done it before, and a "main" one, with 250 /500/ 750 /1000 steps, for people with a taste for real volume. Hopefully, some editors would experiment with on the "starter" one, find they like it and move onto the larger ranges of the "main" worklist. The starter one would use say 10,000 articles blocked off from the second. The instructions would aim to steer people in the right direction.

Is this practical? If not, at least I hope it is clearer, --ROGER DAVIES ^talk 16:34, 7 March 2008 (UTC)

To address an earlier point of yours ("There was considerable support for 1000 article ranges in the workshop"): I disagree. Going by the Comments and Proposals sections from the Workshop, I see at least as much opposition as agreement, although relatively few people gave input.

In RE "starter" lists - I somehow entirely missed that suggestion at the Workshop, and only just now read it. I think that's a great idea for newbies, but for high-volume contributors, I still think the satisfaction and perceived momentum of completing a page is an important consideration.

I promise I'm not going to keep harping on this forever, but I still feel there's a significant benefit to shorter worklist pages. How would you feel about 100x5? Maralia (talk) 17:01, 7 March 2008 (UTC)

• Yes, you're right. It's a few days since I re-read it and I misremembered it.
• "Starter" worklist: 4 x 25-article list. With no graphics, and no overlap into the next editor's range. OK?
• "Main" worklist: 5 x 100 sounds fine. With neither graphics nor overlap as per "starter".
• Please do harp on. It's all constructive stuff :)
• Any other thoughts?

--ROGER DAVIES ^talk 22:49, 7 March 2008 (UTC)

Just a couple suggestions from the workshop:

• On the starter worklists, transclude the instructions every subsection; on the main worklists, only once at the top.
• Reconsider the requirement for strikethrough on out-of-scope articles. I don't know how the completed lists are used after the drive, so it's hard to suggest how to address this, but several people (myself included) found strikethrough really time consuming.

Maralia (talk) 18:23, 10 March 2008 (UTC)

A new approach?

Based on early experiences with BCAD, we might re-consider our approach to this drive. One of the problems is users adopting ranges and then doing little or nothing. One solution to this presents itself though it increases the workload of the people looking after the drive. Here's how it might work:

Let's assume that the summer T&A drive has 100,000 articles. We start by running the script to create the master list. We then post one small worksheet, say the first 5000 articles, but with just say 25 articles in each range. When it is nearly full up, add a second worksheet for the second 5000. When that fills, add a third and remove the first. If the worksheets are transclusions, this is actually easy to do. What is do need is a slightly more complex participant tally sheet, with one column per worksheet. What do we think? --ROGER DAVIES ^talk 22:59, 12 March 2008 (UTC)

That could work, but it may give the impression of a neverending list, which may have the opposite effect and turn people off to the drive. Question: What about splitting the lists up roughly into those regions/periods/topics covered by the task forces? In this manner the parent project can function as the refs of an american football team, letting task forces play the role of teams. This should split the article payload up into somewhat more managable lists, and may attract more participants by catering more to the strengths our of countributers. To be fair, I must state that I have no idea if this is possible or not from a technical stand point. —Preceding unsigned comment added by TomStar81 (talk • contribs)

It's not really possible, unfortunately, as there's no good way to tie categories to task forces externally. Even if it were, you'd wind up with large numbers of overlaps, since the task force scopes aren't mutually exclusive; and that would mean that the total number of list entries would be far greater than the actual number of articles in the list. Kirill 15:37, 16 March 2008 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.